Querying data on the Semantic Web

Observing example queries

In this part, you simply observe the existing examples from Wikidata Query Service.

• Use the button labelled "Examples" (or "Exemples" if your browser is in French).
• Start looking at the first example, listing notable cats. Note that the SERVICE clause here uses special features of Wikidata that are not available in standard SPARQL endpoints. To execute the query, click on the blue button with a white triangle on the left side.
• Put your mouse on top of wdt:P31. What do you read? Do the same for wd:Q146. Write your answers in the file.
• Delete Q146 and place your cursor after wd:. Type Ctrl+Space. Then you should have a little popup that says "Type to search for an entity". Type "cow" and select the entity that corresponds to a kind of cattle and press enter. What is the Q number of this entity? Write your answer in the file.
• Execute the query. How many results there are? Write your answer in the file.
• Select the example query for horses (with some info about them). This query shows multiple SPARQL constructs. Try to understand them.
• The example query named "Cats, with pictures" (or "Chats, avec images") uses a special feature of Wikidata that is not part of the language SPARQL. At the begining of the query, there is comment saying #defaultView:ImageGrid. This turns any URL of pictures in the result of a query into a displayed image.
• The example query "Map of hospitals" (or equivalent in French) uses another feature of Wikidata. At the beggining of the query, there is comment saying #defaultView:Map. Execute the query and see what happens. Then delete the comment and execute again. What allows Wikidata to determine the location of a result? Write your answer in the file.
• Take a close look at the query "Items about authors with a Wikispecies page" (or equivalent in French). What SPARQL constructs are used here that are also SQL constructs? Write your answer in the file.
• The query "Recent events" shows other features, like FILTER, BOUND, comparators, operators, DATATYPE, BIND ... AS, LANG.
• You can play a bit more with the example queries until you are ready to move on to the next part.

Write your own queries

We will write query on Wikidata query Service.

• Go to the Wikidata Query Service.
• The query SELECT * WHERE {?x rdfs:label "Saint-Étienne"@fr .} will give you results that identify entities named "Saint-Étienne" in French. Which result corresponds to the French city? Write the full IRI that identifies the city of Saint-Étienne. What are the other results about?
• Wikidata does not use the standard rdf:type but has a property wdt:P31 (instance of) that more or less plays this role. Write a SPARQL query to find the types of Saint-Étienne. Among the types, you should find wd:Q484170. This corresponds to the "communes in France", the administrative geographic area associated with a city, town or village in France.
• We can get more types for Saint-Étienne by using the class hierarchy. Wikidata does not use the standard rdfs:subClassOf but has something similar: wdt:P279 (subclass of). If we use a property path wdt:P31/wdt:P279*, we can get all of the types and supertypes. How many types does Saint-Étienne has in Wikidata? Write the query and your answer in the file.
• What are the communes existing in Wikidata? (in RDF terms, this means "things of type commune", assuming there is an appropriate IRI for the class of communes). Write your query in the file.
• List the communes with their population (hint: use the autocompletion feature after write wdt: using Ctrl+Space). You can also find it by looking at a well known city in Wikidata, e.g., Saint-Étienne). Write your query in the file.
• List communes that have more than 100,000 inhabitants, ordered by population. Write your query in the file.
• List countries with their capitals. Write your query in the file.
• List countries ordered by the population of their capitals. Write your query in the file.
• Find the total population of communes in région Rhône-Alpes. Write your query in the file.
• Find communes in region Rhône-Alpes and, if available, a picture of it. Write your query in the file.

More challenges

You may try to find who are the Physics Nobel prize laureates from Europe.

You may also try to find who are the widow of assassinated USA presidents. Then of all presidents, regardless of the country.

Some useful references

• RDF Playground from Univeristy of Chile, with RDF syntax checker and visualiser, SPARQL editor and engine, and more things.
• SPARQL Playground from the Swiss institute of bioinformatics, with more and more complicated examples.